Programmers at the time were okay with this and it was actually fun. The tools though were much better for the competition, on PS2 I could set break points anywhere (even on memory reads) and see the DMA streams in the debugger, on NUON it was pretty much debugging with prints except you had to write the print yourself :)
A gaming system (which this was) can't wait for that popularity to happen through some other means. The games are the raison d'être of the system, the only reason for people to buy one. What was the end-user pitch here that'd get one of these to the hands of millions of consumers? "You can buy some other DVD player for $200, or this one that might some day play games for $300". Not very compelling, when both of them do an equally good job with the main task you're buying a DVD player for.
I wouldn't say this is an inevitability. Look at Nokia feature-phones.
Sure, from what I've heard and read an incredibly poorly designed ecosystem and lack of carrier interest, but you can't tell me Snake was the best the market could come up with...
The dev team got paid for ~7 years?
It's really funny, because all throughout reading that article I was thinking "this thing never had a chance". A custom architecture based on principles that are notoriously difficult to write tools for (VLIW), probably adding 50% to the BOM of very price sensitive consumer electronics, and an attempt to create a gaming platform by a company with no capability of producing first party titles and no commitments from credible publishers to do third party games either. How could that not fail?
But then the article ends with all of these people believing - even with 15 years of hindsight - that they genuinely were just on the verge of breaking through. Their technology was just too far ahead of its time and people didn't yet understand it!